Use of the SINBAD Classification System and Score in Comparing Outcome of Foot Ulcer Management on Three Continents

The contribution to outcome made by demographic features and baseline ulcer characteristics has been reported in a consecutive cohort of 449 patients referred to a specialist clinic in Nottingham, U.K., over 4 years (8). Each lesion was classified using the S(AD)SAD system at the time of first attendance and recategorized retrospectively for the purposes of the present study using the SINBAD system (Table 1). The data to define the site were available from the clinic database. In the SINBAD system, these six elements were graded as follows: 1) ulcer site (forefoot [distal to tarso-metatarsal joint], 0, and midfoot/hindfoot, 1); 2) ischemia (blood flow relatively intact [at least one pulse palpable on the affected foot], 0, and evidence of ischemia [neither pulse palpable with signs of reduced tissue perfusion, with or without gangrene], 1); 3) neuropathy, defined as being absent, 0, or present, 1, on the basis of routine examination using either Neurotips (Owen Mumford) or 10-g nylon monofilaments; 4) bacterial infection (using clinical signs of infection of either soft tissue or bone proposed by the Infectious Diseases Society of America and the International Working Group on the Diabetic Foot) (16,17), graded as absent, 0, or present, 1; 5) area (the two maximum dimensions at right angles multiplied [≤1 cm², 0, and >1 cm², 1]; and 6) depth (superficial, 0, and deep [reaching to tendon, periosteum, joint capsule, or bone], 1) (Table 1). In addition, the individual grades were summed, creating a SINBAD score range of 0–6. We sought to identify the contribution made by each of the six elements to outcome (healed vs. nonhealed) by using χ² and logistic regression. An association was also sought between SINBAD score and median time to healing. These analyses were repeated using data collected prospectively in consecutive series of patients managed in specialist clinics in Dar es Salaam (Tanzania), Karachi (Pakistan), and Soest (Germany). Foot care services in Pakistan are private, whereas those in the other countries are free at the point of delivery and paid for by either central government or personal insurance. The data from the patients managed in Dar es Salaam have previously been used to compare four different systems of classification (Wagner, University of Texas, S(AD)SAD, and PEDIS [Perfusion, Extent, Depth, Infection, and Sensation]) (18). The data from Germany have been compared with those of separate populations managed in Tanzania and India (19). Univariate analysis was undertaken seeking differences between baseline variables and outcome (healed vs. nonhealed, including amputation and death) using Pearson's χ² or Fisher's exact tests. Variables with a significance of P < 0.05 were entered for multivariate analysis by logistic regression. It has previously been established by the Caldicott Guardian in Nottingham that ethical approval is not required to analyze anonymous data collected during the course of routine management.

Data from the 449 consecutive cases managed in Nottingham were used to analyze outcomes according to the simplified baseline categories and to explore the significance of the aggregate SINBAD score. The median time to healing in the 323 cases that healed without amputation was determined for ulcers with different SINBAD scores, and a highly significant difference between groups was found (χ² 37.324, Kruskal-Wallis test; P = 0). Data from the other three centers were then classified according to the new system.

Table 2 summarizes baseline demographic details and outcomes of the ulcers managed at the four centers. Outcomes were determined at a fixed date in each center; median time to outcome varied from 30 to 91 days with maximum durations of follow-up ranging from 973 to 1,344 days. The prevalence of ulcers of different SINBAD score in the four centers is shown in Table 3 and the median time to healing (of ulcers that healed without amputation) in Table 4.

Univariate analysis revealed significant differences between the baseline categories associated with outcome (healing vs. nonhealing) in the four centers (Table 5). Ulcer depth was the only variable associated with healing in the series from Karachi, although trends to differences were observed for site (Fisher's exact test; P = 0.069) and for ischemia (Fisher's exact test; P = 0.090). The results of multivariate analysis are shown in Table 6.

This study reveals that different baseline ulcer characteristics are associated with outcome in different countries. It also shows that these characteristics can be expressed in an aggregate SINBAD score, with a score of three or more being associated with a step-up in time to healing and in eventual failure to heal. The SINBAD score may therefore represent a system for defining ulcer type that could be applied worldwide. A strength of the study is that it employs data collected prospectively in consecutive series of patients. A limitation lies in the outcome data being based on variable duration of follow-up, weakening to some extent the descriptive analysis of eventual outcome (healing, nonhealing, amputation, and death).

The results confirm the marked differences in the nature of foot lesions managed in different countries and in patient age, as previously reported (19). Those in less industrialized countries are ∼15 years younger and because there is also a lower prevalence of peripheral arterial disease, foot ulcers in developing countries tend to be caused by trauma from inappropriate or nonexistent footwear and are frequently complicated by infection. When access to free health care is not available (as in Pakistan) or delayed because of initial reliance on traditional healers (as in Tanzania), patients may present late and with more severe infection. It has been estimated that the management of a single ulcer costs on average between 0.3 and 5.5 times the average total monthly expenditure for a household in Pakistan (A.B., S.M.A., A. Fareed, S. Mujahid Humail, M. Yakoob Ahmedani, A. Fawwad, Z. Miyan, unpublished data). It is probably for such reasons that in the series from these two countries, relatively uncomplicated lesions (those with SINBAD scores of 0–2, Table 3) were underrepresented, whereas the number lost to follow-up was high.

The differences in patient population and ulcer type emphasize the need for using appropriate systems of classification to compare ulcer outcomes among centers. Such systems must encompass all the variables that may be associated with outcome, and our findings reveal that these vary considerably, with only depth being common to all centers on univariate analysis (Table 5). By using an aggregate scoring system, however, it is possible to obtain a measure of ulcer type independent of the specific factors of which it is comprised and enables such comparisons to be made. Inspection of Table 4 reveals a clear step-up in time to healing between SINBAD scores of 2 and 3 in each of the four centers, suggesting that those with ulcers of grade 3 or greater are at particular risk.

Interesting differences were also observed between outcomes in the studied populations in the U.K. and Germany. The main difference between the two lies in the strong association that exists in Germany between infection and outcome; the association was only weak in the U.K. and lost in multivariate analysis. The incidence of amputation (major and minor combined) was also much higher in the German cohort (19.2 vs. 8.0%), although it is important to note that the two series were not concurrent, dating from 1998 to 1999 and 2000 to 2003, respectively, and the incidence of amputation for all new ulcers in Soest has since fallen by 60% (S.M., unpublished data). From 1998 to 1999, however, it is possible that it was more common in Germany to manage osteomyelitis by early elective excision of bone, whereas most centers in the U.K. reserve amputation for those with either overwhelming infection or who fail to respond to initial antibiotic treatment (19,20). In this respect, the findings in Germany resembled those in North America, where a similarly close association has been shown between outcome (incidence of amputation) and infection at baseline (5) and where early surgery is more common practice. Conversely, some 11.6% of the population managed in the U.K. had persisting disease at the time of assessment in the present series versus 0% in Germany. Moreover, the time to healing was longer in the U.K. than in any of the other three centers for ulcers of all grades and, although this could reflect less effective care, it could also result indirectly from lower incidence of amputation and an attempt to achieve healing without surgery in a greater proportion of the population. Detection of differences such as these highlights the value of comparative audit and pinpoints the need for formal assessment of the advantages and disadvantages of different strategies of care.

We thank Shabneez Gangji and Kulsum Ramadhani in Dar es Salaam and the clinical staff at each of our specialist centers.